Understanding mechanisms of Transcription Factor cooperativity across scales

Introduction

Transcription Factors (TFs) are critical regulators of many essential cellular functions such as the acquisition of cell identities in healthy tissues and their dysregulation in disease. Transcriptional activation of a gene typically requires the cooperative binding of multiple TFs, which subsequently recruit various additional cofactors.

Background

Genomics has enabled the generation of a near-complete annotation of the cis-regulatory elements and TFs binding them across cell types. Yet, the precise function of each TF in the process and how these functionalities are assembled to activate transcription is an important open question.

Hypothesis

Here we postulate that despite strong cell-type specificity, the formation of TF cooperativity modules on DNA relies on general principles that are shared across cell types.

Project Proposal

In TFCoop, we propose to formalize these organizational rules by probing the effect of hundreds of thousands of perturbations of individual TFs on the regulatory activity of their network.

Methodology

1. We will apply time-resolved nuclear depletion using optogenetics in parallel for multiple TFs of two related networks.
2. We will contrast the primary effects of their depletion genome-wide.
3. In a complementary approach, we will develop a reductionist system to study the function of tens of thousands of individual or controlled combinations of TF motifs when inserted into the genome.

Techniques

We will leverage the unique properties of single molecule genomics to measure the contribution of each TF to the activity of multiple components of the regulatory system across multiple loci simultaneously.

Data Analysis

This will be followed by factor analysis and deep learning to integrate this large collection of primary effects of TF perturbation and identify the general principles of their assembly into cooperativity networks.

Expected Outcomes

Upon the success of the project, the resulting models will unlock the understanding of the genetic encoding of cellular identities and allow their manipulation for regenerative medicine.